NeighborMom() vs AnyMom()

The way a mother is chosen after death has impacts on the result of the simulation. Below are two methods for selecting that individual to reproduce: weighted sampling by fitness of all individuals (anyMom), or weighted sampling by fitness of only individuals adjacent to the dead one (neighborMom).

function neighborMom(network::NetworkParameters, kID::Int64)
    fitWeights = StatsBase.weights([network.popFitness[kID%100+1], network.popFitness[(kID+98)%100+1]]) #index plus/minus one, loops around 100
    momIndex = (kID + sample([1, -1], fitWeights) + 99)%100 + 1 #have to use (idx +99)%100 + 1 to loop from 100 to 1 properly
    momIndex
end

function anyMom(network::NetworkParameters, kID::Int64)
    network.popFitness[kID] = 0
    fitWeights = StatsBase.weights(network.popFitness)
    momIndex = sample(1:network.popSize, fitWeights)
    resolveLocs(network, kID, momIndex) #updates locations, ensures each individual has a unique location by inserting the child adjacent to the parent and pushing others away 
    momIndex
end

UndefVarError: NetworkParameters not defined



Stacktrace:

 [1] top-level scope

   @ c:\Users\jonah\OneDrive\Desktop\Research\Parker Model\neighbor-mom.ipynb:1

Both NeighborMom() and AnyMom() serve as different methods replacing FindMom() based on our method of selecting parents. NeighborMom() only compares fitnesses of neighboring individuals, while AnyMom() looks at fitnesses of all individuals in the population. Locations do not need to be separately updated for NeighborMom(), as each individual is automatically adjacent to a parent.

Old heatmap for Iterating by Ben/CL

using JLD2
using StatsBase
using Plots
#backend(:plotly)
using DataFrames
using CSV

function dataRead(data, range, index::Int)
    df = CSV.read(data, DataFrame, delim=",")
    df = df[!, [2, 3, 5, 6, 8, 13, 14, 15, 16, 17, 4, 11, 10, 12, 7, 9, 1]] #used for soring by ben/cl, including graphCalcs
    #df = df[!, [6, 9, 2, 3, 7, 8, 10, 4, 5, 1]] #used for when sorting by pn/r
    #sort!(df, [:pn, :pr]) #used for when sorting by pn/r
    sort!(df, [:ben, :cl])
    dfData = Matrix{Float64}(df)
    dataArr = zeros(range, range, 1) #transforms [range^2 * 10] arr to [range, range, 1] arr
    for(i) in 1:range
        for(j) in 1:range
            dataArr[i, j, 1] = dfData[range*(i-1)+j, index]
        end
    end
    #print(dataArr)
    dataArr
end

function hmap(data::String, range::Int, index::Int, titles::String, cLow::Float64, chI::Float64) 
    #indexes : 1 yidx   2 xidx   3 coopFreq   4 degree   5 fitness  6 pnc_end   7 pnd_end   8 prc_end  9 prd_end  10 shortestPath 11 connComponents 12 meanCCSize 13 largestCC 14 mean connection distance 15 distance  16 inclusion  17 assortment 
    dataArr = dataRead(data, range, index)
    x_axis = String[]
    y_axis = String[]
    for(i) in 1:range #PNC/PND times 10
        push!(x_axis, (string(round((0.4*i/range); digits = 4))))
        push!(y_axis, (string(round((10*i/range); digits = 3))))
    end
    p = heatmap(x_axis, y_axis,  dataArr[:, :, 1]; title = titles, xlabel="CL",ylabel="BEN", clim=(cLow,chI))
    p
end

#redefined hmap() in between runs

hmap (generic function with 2 methods)

High Strength of Selection, AnyMom(), no Distance-based inheritance

The following two runs were created before distance inheritance was implemented, so do not have the mean connection distance and have to use an older hmap() function.

file = "anyMom_no_pnd_new.csv"
fig1hmaps = [hmap(file, 20, 3, "Cooperation Frequency", 0.0, 1.0), 
hmap(file, 20, 6, "PN", 0.0, 1.0), 
hmap(file, 20, 8, "PR", 0.0, 0.3),
hmap(file, 20, 10, "Mean Shortest Path", 1.0, 4.0), 
hmap(file, 20, 11, "Connected Components", 1.0, 45.0), 
hmap(file, 20, 12, "Mean Conn. Comp. Size", 1.0, 100.0), 
hmap(file, 20, 13, "Largest Conn. Comp.", 10.0, 100.0),
hmap(file, 20, 4, "Degree", 1.0, 75.0)] 

plot(fig1hmaps..., colorbar = true, grid=false, layout = (4,2), size = (1000, 1200))

Figure 1 High strength of selection with the mother selected from any individual, pnd and prd not evolving separately. The results above were created with: C=0.5, D=0.0, gen=100000, pn=0.5, distInherit=false, pnd=false, pr=0.0001, prd=false, muP=0.01, delta=0.5, sigmapn=0.01, sigmapr=0.01, reps=10. Initial cooperation frequency was 0.5.

file = "anyMom_pnd_new.csv"
fig1hmaps = [hmap(file, 20, 3, "Cooperation Frequency", 0.0, 1.0), 
hmap(file, 20, 6, "PNC", 0.0, 1.0), 
hmap(file, 20, 7, "PND", 0.0, 1.0),
hmap(file, 20, 8, "PR", 0.0, 0.3),
hmap(file, 20, 10, "Mean Shortest Path", 1.0, 4.0), 
hmap(file, 20, 11, "Connected Components", 1.0, 45.0), 
hmap(file, 20, 12, "Mean Conn. Comp. Size", 1.0, 100.0), 
hmap(file, 20, 13, "Largest Conn. Comp.", 10.0, 100.0),
hmap(file, 20, 4, "Degree", 1.0, 75.0)] 

plot(fig1hmaps..., colorbar = true, grid=false, layout = (5,2), size = (1000, 1500))

Figure 2 High strength of selection with the mother selected from any individual, pnd but not prd evolving separately. The results above were created with: C=0.5, D=0.0, gen=100000, pn=0.5, distInherit=false, pnd=true, pr=0.0001, prd=false, muP=0.01, delta=0.5, sigmapn=0.01, sigmapr=0.01, reps=10. Initial cooperation frequency was 0.5.

New Heatmap Function

function dataRead(data, range, index::Int)
    df = CSV.read(data, DataFrame, delim=",")
    df = df[!, [2, 3, 5, 6, 8, 13, 14, 15, 16, 17, 4, 11, 10, 12, 7, 9, 1]] #used for soring by ben/cl, including graphCalcs
    #df = df[!, [6, 9, 2, 3, 7, 8, 10, 4, 5, 1]] #used for when sorting by pn/r
    #sort!(df, [:pn, :pr]) #used for when sorting by pn/r
    sort!(df, [:ben, :cl])
    dfData = Matrix{Float64}(df)
    dataArr = zeros(range, range, 1) #transforms [range^2 * 10] arr to [range, range, 1] arr
    for(i) in 1:range
        for(j) in 1:range
            dataArr[i, j, 1] = dfData[range*(i-1)+j, index]
        end
    end
    #print(dataArr)
    dataArr
end

function hmap(data::String, range::Int, index::Int, titles::String, cLow::Float64, chI::Float64) 
    #indexes : 1 yidx   2 xidx   3 coopFreq   4 degree   5 fitness  6 pnc_end   7 pnd_end   8 prc_end  9 prd_end  10 shortestPath 11 connComponents 12 meanCCSize 13 largestCC 14 mean connection distance 15 distance  16 inclusion  17 assortment 
    dataArr = dataRead(data, range, index)
    x_axis = String[]
    y_axis = String[]
    for(i) in 1:range #PNC/PND times 10
        push!(x_axis, (string(round((0.4*i/range); digits = 4))))
        push!(y_axis, (string(round((10*i/range); digits = 3))))
    end
    p = heatmap(x_axis, y_axis,  dataArr[:, :, 1]; title = titles, xlabel="CL",ylabel="BEN", clim=(cLow,chI))
    p
end

#redefined hmap() in between runs

hmap (generic function with 2 methods)

High Strength of Selection, NeighborMom(), no Distance-based inheritance

file = "neighbor_mom_no_pnd_fixed.csv"
fig1hmaps = [hmap(file, 20, 3, "Cooperation Frequency", 0.0, 1.0), 
hmap(file, 20, 6, "PN", 0.0, 0.7), 
hmap(file, 20, 8, "PR", 0.0, 0.40),
hmap(file, 20, 14, "Mean Dist. of Connections", 1.0, 30.0),
hmap(file, 20, 10, "Mean Shortest Path", 1.0, 4.25), 
hmap(file, 20, 11, "Connected Components", 1.0, 35.0), 
hmap(file, 20, 12, "Mean Conn. Comp. Size", 1.0, 100.0), 
hmap(file, 20, 13, "Largest Conn. Comp.", 1.0, 100.0),
hmap(file, 20, 4, "Degree", 1.0, 75.0)
]
#,

plot(fig1hmaps..., colorbar = true, grid=false, layout = (5,2), size = (1000, 1500))

Figure 3 High strength of selection with the mother selected from dead individuals' neighbors, pnd and prd not evolving separately. The results above were created with: C=0.5, D=0.0, gen=100000, distInherit=false, distFactor=1.0, pn=0.5, findMom=neighborMom, pnd=false, pr=0.0001, prd=false, muP=0.01, delta=0.5, sigmapn=0.01, sigmapr=0.01, reps=10). Initial cooperation frequency was 0.5 .

file = "neighbor_mom_pnd.csv"
fig1hmaps = [hmap(file, 20, 3, "Cooperation Frequency", 0.0, 1.0), 
hmap(file, 20, 6, "PNC", 0.0, 1.0), 
hmap(file, 20, 7, "PND", 0.0, 1.0),
hmap(file, 20, 8, "PR", 0.0, 0.50),
hmap(file, 20, 14, "Mean Dist. of Connections", 1.0, 30.0),
hmap(file, 20, 10, "Mean Shortest Path", 1.0, 4.25), 
hmap(file, 20, 11, "Connected Components", 1.0, 30.0), 
hmap(file, 20, 12, "Mean Conn. Comp. Size", 1.0, 100.0), 
hmap(file, 20, 13, "Largest Conn. Comp.", 1.0, 100.0),
hmap(file, 20, 4, "Degree", 1.0, 100.0)
]
#,

plot(fig1hmaps..., colorbar = true, grid=false, layout = (5,2), size = (1000, 1500))

Figure 4 High strength of selection with the mother selected from dead individuals' neighbors, pnd but not prd evolving separately. The results above were created with: C=0.5, D=0.0, gen=100000, distInherit=false, distFactor=1.0, pn=0.5, findMom=neighborMom, pnd=true, pr=0.0001, prd=false, muP=0.01, delta=0.5, sigmapn=0.01, sigmapr=0.01, reps=10. Initial cooperation frequency was 0.5 .

High Strength of Selection, NeighborMom(), Distance-based inheritance

file = "neighbor_mom_no_pnd_0.95_dist.csv"
fig1hmaps = [hmap(file, 20, 3, "Cooperation Frequency", 0.0, 1.0), 
hmap(file, 20, 6, "PN", 0.0, 0.7), 
hmap(file, 20, 8, "PR", 0.0, 0.35),
hmap(file, 20, 14, "Mean Dist. of Connections", 1.0, 40.0),
hmap(file, 20, 10, "Mean Shortest Path", 1.0, 3.5), 
hmap(file, 20, 11, "Connected Components", 1.0, 40.0), 
hmap(file, 20, 12, "Mean Conn. Comp. Size", 1.0, 100.0), 
hmap(file, 20, 13, "Largest Conn. Comp.", 1.0, 100.0),
hmap(file, 20, 4, "Degree", 1.0, 100.0)
]
#,

plot(fig1hmaps..., colorbar = true, grid=false, layout = (5,2), size = (1000, 1500))

Figure 5 High strength of selection with the mother selected from dead individuals' neighbors, neither pnd nor prd evolving separately. The results above were created with: C=0.5, D=0.0, gen=100000, distInherit=false, distFactor=1.0, pn=0.5, findMom=neighborMom, pnd=false, pr=0.0001, prd=false, muP=0.01, delta=0.5, sigmapn=0.01, sigmapr=0.01, reps=10. Initial cooperation frequency was 0.5 .

file = "neighbor_mom_pnd_0.95_dist.csv"
fig1hmaps = [hmap(file, 20, 3, "Cooperation Frequency", 0.0, 1.0), 
hmap(file, 20, 6, "PNC", 0.0, 0.9), 
hmap(file, 20, 7, "PND", 0.0, 0.9),
hmap(file, 20, 8, "PR", 0.0, 0.35),
hmap(file, 20, 14, "Mean Dist. of Connections", 1.0, 35.0),
hmap(file, 20, 10, "Mean Shortest Path", 1.0, 3.0), 
hmap(file, 20, 11, "Connected Components", 1.0, 40.0), 
hmap(file, 20, 12, "Mean Conn. Comp. Size", 1.0, 100.0), 
hmap(file, 20, 13, "Largest Conn. Comp.", 1.0, 100.0),
hmap(file, 20, 4, "Degree", 1.0, 100.0)
]
#,

plot(fig1hmaps..., colorbar = true, grid=false, layout = (5,2), size = (1000, 1500))

Figure 6 High strength of selection with the mother selected from dead individuals' neighbors, pnd but not prd evolving separately. The results above were created with: C=0.5, D=0.0, gen=100000, distInherit=false, distFactor=1.0, pn=0.5, findMom=neighborMom, pnd=true, pr=0.0001, prd=false, muP=0.01, delta=0.5, sigmapn=0.01, sigmapr=0.01, reps=10. Initial cooperation frequency was 0.5 .

Changing DistFactor / Cost of Linking

Heatmap for Iterating by DistFactor/CL

function dataRead(data, range, index::Int)
    df = CSV.read(data, DataFrame, delim=",")
    df = df[!, [2, 6, 4, 5, 8, 13, 14, 15, 16, 17, 3, 11, 10, 12, 7, 9, 1]] #used for soring by clratio/distFactor, including graphCalcs
    #df = df[!, [2, 3, 5, 6, 8, 13, 14, 15, 16, 17, 4, 11, 10, 12, 7, 9, 1]] #used for soring by ben/cl, including graphCalcs
    #sort!(df, [:pn, :pr]) #used for when sorting by pn/r
    sort!(df, [:cl , :distFactor])
    #print(df)
    dfData = Matrix{Float64}(df)
    dataArr = zeros(range, range, 1) #transforms [range^2 * 10] arr to [range, range, 1] arr
    for(i) in 1:range
        for(j) in 1:range
            dataArr[i, j, 1] = dfData[range*(i-1)+j, index]
        end
    end
    #print(dataArr)
    dataArr
end

function hmap(data::String, range::Int, index::Int, titles::String, cLow::Float64, chI::Float64) 
    #indexes : 1 yidx   2 xidx   3 coopFreq   4 degree   5 fitness  6 pnc_end   7 pnd_end   8 prc_end  9 prd_end  10 shortestPath 11 connComponents 12 meanCCSize 13 largestCC 14 mean connection distance 15 distance  16 inclusion  17 assortment 
    dataArr = dataRead(data, range, index)
    x_axis = String[]
    y_axis = String[]
    for(i) in 1:range 
        push!(x_axis, (string(round(0.5+0.5*i/range; digits = 2))))
        push!(y_axis, (string(round(0.4*i/range; digits = 3))))
    end
    p = heatmap(x_axis, y_axis,  dataArr[:, :, 1]; title = titles, ylabel="CL",xlabel="DistFactor", clim=(cLow,chI))
    p
end

hmap (generic function with 2 methods)

file = "anymom_dist_fixed.csv"
fig1hmaps = [hmap(file, 20, 3, "Cooperation Frequency", 0.0, 1.0), 
hmap(file, 20, 6, "PNC", 0.0, 0.8), 
hmap(file, 20, 7, "PND", 0.0, 0.8),
hmap(file, 20, 8, "PR", 0.0, 0.6),
hmap(file, 20, 14, "Mean Dist. of Connections", 1.0, 28.0),
hmap(file, 20, 10, "Mean Shortest Path", 1.0, 6.0), 
hmap(file, 20, 11, "Connected Components", 1.0, 40.0), 
hmap(file, 20, 12, "Mean Conn. Comp. Size", 1.0, 100.0), 
hmap(file, 20, 13, "Largest Conn. Comp.", 1.0, 100.0),
hmap(file, 20, 4, "Degree", 1.0, 35.0)
]

plot(fig1hmaps..., colorbar = true, grid=false, layout = (5,2), size = (1000, 1500))

Figure 7 High strength of selection with the mother selected from any individual, distance-based inheritance enabled, pnd but not prd evolving separately. The results above were created with: B=2.0, C=0.5, D=0.0, gen=100000, distInherit=true, findMom=anyMom, pn=0.5, pnd=true, pr=0.0001, prd=false, muP=0.01, delta=0.5, sigmapn=0.01, sigmapr=0.01, reps=10. Initial cooperation frequency was 0.5.

file = "nmom_dist_fixed.csv"
fig1hmaps = [hmap(file, 20, 3, "Cooperation Frequency", 0.0, 1.0), 
hmap(file, 20, 6, "PNC", 0.0, 0.8), 
hmap(file, 20, 7, "PND", 0.0, 0.8),
hmap(file, 20, 8, "PR", 0.0, 0.6),
hmap(file, 20, 14, "Mean Dist. of Connections", 1.0, 28.0),
hmap(file, 20, 10, "Mean Shortest Path", 1.0, 6.0), 
hmap(file, 20, 11, "Connected Components", 1.0, 40.0), 
hmap(file, 20, 12, "Mean Conn. Comp. Size", 1.0, 100.0), 
hmap(file, 20, 13, "Largest Conn. Comp.", 1.0, 100.0),
hmap(file, 20, 4, "Degree", 1.0, 35.0)
]

plot(fig1hmaps..., colorbar = true, grid=false, layout = (5,2), size = (1000, 1500))

Figure 8 High strength of selection with the mother selected from dead individuals' neighbors, distance-based inheritance enabled, pnd but not prd evolving separately. The results above were created with: B=2.0, C=0.5, D=0.0, gen=100000, distInherit=true, findMom=neighborMom, pn=0.5, pnd=true, pr=0.0001, prd=false, muP=0.01, delta=0.5, sigmapn=0.01, sigmapr=0.01, reps=10. Initial cooperation frequency was 0.5.

NeighborRange

The following introduces a new parameter, neighborRange, and modifies neighborMom() as below to allow for an input range to control how many individuals are possible to be selected for the mother.

function neighborMom(network::NetworkParameters, kID::Int64, neighborRange::Int64=1)
    indexes = zeros(Int64, 2*neighborRange)
    fitnesses = zeros(2*neighborRange)
    for(i) in 1:neighborRange
        indexes[i] = (kID-i+99)%100+1 #index minus 1 through neighborRange, loops around 100
        indexes[i+neighborRange] = (kID+i+99)%100+1 #index plus 1 through neighborRange, loops around 100
        fitnesses[i] = network.popFitness[indexes[i]] 
        fitnesses[i+neighborRange] = network.popFitness[indexes[i+neighborRange]]
    end
    fitWeights = StatsBase.weights(fitnesses)
    return sample(indexes, fitWeights) 
end

UndefVarError: NetworkParameters not defined



Stacktrace:

 [1] top-level scope

   @ c:\Users\jonah\OneDrive\Desktop\Research\Parker Model\neighbor-mom.ipynb:1

Heatmap for iterating by neighborRange (1-20) and cl

function dataRead(data, range, index::Int)
    df = CSV.read(data, DataFrame, delim=",")
    df = df[!, [2, 6, 4, 5, 8, 13, 14, 15, 16, 17, 3, 11, 10, 12, 7, 9, 1]] #used for soring by clratio/distFactor/distNeighborRange, including graphCalcs
    sort!(df, [:distNeighborRange, :cl])
    dfData = Matrix{Float64}(df)
    dataArr = zeros(range, range, 1) #transforms [range^2 * 10] arr to [range, range, 1] arr
    for(i) in 1:range
        for(j) in 1:range
            dataArr[i, j, 1] = dfData[range*(i-1)+j, index]
        end
    end
    dataArr
end

function hmap(data::String, range::Int, index::Int, titles::String, cLow::Float64, chI::Float64) 
    #indexes : 1 yidx   2 xidx   3 coopFreq   4 degree   5 fitness  6 pnc_end   7 pnd_end   8 prc_end  9 prd_end  10 shortestPath 11 connComponents 12 meanCCSize 13 largestCC 14 mean connection distance 15 distance  16 inclusion  17 assortment 
    dataArr = dataRead(data, range, index)
    x_axis = String[]
    y_axis = String[]
    for(i) in 1:range 
        push!(y_axis, (string(Int(20*i/range))))
        push!(x_axis, (string(round(0.4*i/range; digits = 3))))
    end
    p = heatmap(x_axis, y_axis,  dataArr[:, :, 1]; title = titles, xlabel="CL",ylabel="Neighbor Range", clim=(cLow,chI))
    p
end

hmap (generic function with 2 methods)

file = "neighborRange-1-20.csv"
fig1hmaps = [hmap(file, 20, 3, "Cooperation Frequency", 0.0, 1.0), 
hmap(file, 20, 6, "PNC", 0.0, 0.8), 
hmap(file, 20, 7, "PND", 0.0, 0.8),
hmap(file, 20, 8, "PR", 0.0, 0.6),
hmap(file, 20, 14, "Mean Dist. of Connections", 1.0, 28.0),
hmap(file, 20, 10, "Mean Shortest Path", 1.0, 6.0), 
hmap(file, 20, 11, "Connected Components", 1.0, 40.0), 
hmap(file, 20, 12, "Mean Conn. Comp. Size", 1.0, 100.0), 
hmap(file, 20, 13, "Largest Conn. Comp.", 1.0, 100.0),
hmap(file, 20, 4, "Degree", 1.0, 35.0)
]

plot(fig1hmaps..., colorbar = true, grid=false, layout = (5,2), size = (1000, 1500))

Figure 9 High strength of selection with the mother selected from individuals a distance neighborRange or below from , pnd but not prd evolving separately. The results above were created with: B=2.0, C=0.5, D=0.0, gen=100000, distInherit=false, findMom=neighborMom, pn=0.5, pnd=true, pr=0.0001, prd=false, muP=0.01, delta=0.5, sigmapn=0.01, sigmapr=0.01, reps=10. Initial cooperation frequency was 0.5.

Finding the difference from both runs with anyMom() and neighborMom() and Figure 9. New heatmap below.

using JLD2
using StatsBase
using Plots
#backend(:plotly)
using DataFrames
using CSV

function benClDataRead(data, range, index::Int)
    df = CSV.read(data, DataFrame, delim=",")
    df = df[!, [3, 2, 5, 6, 8, 13, 14, 15, 16, 17, 4, 11, 10, 12, 7, 9, 1]] #used for soring by ben/cl, including graphCalcs
    #df = df[!, [6, 9, 2, 3, 7, 8, 10, 4, 5, 1]] #used for when sorting by pn/r
    #sort!(df, [:pn, :pr]) #used for when sorting by pn/r
    sort!(df, [:cl, :ben])
    dfData = Matrix{Float64}(df)
    dataArr = zeros(range, range, 1) #transforms [range^2 * 10] arr to [range, range, 1] arr
    for(i) in 1:range
        for(j) in 1:range
            dataArr[i, j, 1] = dfData[range*(i-1)+j, index]
        end
    end
    #print(dataArr)
    dataArr
end

function dataReadMinus(data, data2, range, index::Int)
    df = CSV.read(data, DataFrame, delim=",")
    df = df[!, [2, 6, 4, 5, 8, 13, 14, 15, 16, 17, 3, 11, 10, 12, 7, 9, 1]] #used for soring by clratio/distFactor/distNeighborRange, including graphCalcs
    sort!(df, [:cl, :distNeighborRange])
    dfData = Matrix{Float64}(df)
    #print(df)
    dataArr = zeros(range, range, 1) #transforms [range^2 * 10] arr to [range, range, 1] arr
    for(i) in 1:range
        for(j) in 1:range
            dataArr[i, j, 1] = dfData[range*(i-1)+j, index]
        end
    end
    benclData = benClDataRead(data2, range, index)
    for(i) in 1:range
        dataArr[i, :, 1] .-= benclData[i, 4, 1]
        #dataArr .-= benclData[:, 4, 1]
        #println(dataArr[i, :, 1], " - ", benclData[i, 4, 1])
    end   
    dataArr
end

function hmap(data::String, data2::String, range::Int, index::Int, titles::String, cLow::Float64, chI::Float64) 
    #indexes : 1 yidx   2 xidx   3 coopFreq   4 degree   5 fitness  6 pnc_end   7 pnd_end   8 prc_end  9 prd_end  10 shortestPath 11 connComponents 12 meanCCSize 13 largestCC 14 mean connection distance 15 distance  16 inclusion  17 assortment 
    dataArr = dataReadMinus(data, data2, range, index)
    x_axis = String[]
    y_axis = String[]
    for(i) in 1:range 
        push!(y_axis, (string(Int(20*i/range))))
        push!(x_axis, (string(round(0.4*i/range; digits = 3))))
    end
    p = heatmap(x_axis, y_axis,  dataArr[:, :, 1]; title = titles, xlabel="CL",ylabel="Neighbor Range", clim=(cLow,chI))
    p
end

hmap (generic function with 2 methods)

newdata = "neighborRange-1-20.csv"
compData = "anyMom_pnd_new.csv"

fig1hmaps = [hmap(newdata, compData, 20, 3, "Δ Cooperation Frequency", -0.25, 0.15), 
hmap(newdata, compData, 20, 6, "Δ PNC", -0.3, 0.3), 
hmap(newdata, compData, 20, 7, "Δ PND", -0.3, 0.3),
hmap(newdata, compData, 20, 8, "Δ PR", -0.1, 0.05),
hmap(newdata, compData, 20, 14, "Δ Mean Dist. of Connections", -12.5, 12.5),
hmap(newdata, compData, 20, 10, "Δ Mean Shortest Path", -0.5, 2.0), 
hmap(newdata, compData, 20, 11, "Δ Connected Components", -10.0, 10.0), 
hmap(newdata, compData, 20, 12, "Δ Mean Conn. Comp. Size", -40.0, 30.0), 
hmap(newdata, compData, 20, 13, "Δ Largest Conn. Comp.", -25.0, 5.0),
hmap(newdata, compData, 20, 4, "Δ Degree", -5.0, 5.0)
]

plot(fig1hmaps..., colorbar = true, grid=false, layout = (5,2), size = (1000, 1500))

Figure 10 Difference between Figure 9 (with neighborRange) and Figure 2 (anyMom) at same inputs, besides findMom method

newdata = "neighborRange-1-20.csv"
compData = "neighbor_mom_pnd.csv"

fig1hmaps = [hmap(newdata, compData, 20, 3, "Δ Cooperation Frequency", -0.15, 0.20), 
hmap(newdata, compData, 20, 6, "Δ PNC", -0.3, 0.3), 
hmap(newdata, compData, 20, 7, "Δ PND", -0.3, 0.3),
hmap(newdata, compData, 20, 8, "Δ PR", -0.05, 0.1),
hmap(newdata, compData, 20, 14, "Δ Mean Dist. of Connections", -5.0, 12.5),
hmap(newdata, compData, 20, 10, "Δ Mean Shortest Path", -2.0, 0.5), 
hmap(newdata, compData, 20, 11, "Δ Connected Components", -10.0, 10.0), 
hmap(newdata, compData, 20, 12, "Δ Mean Conn. Comp. Size", -10.0, 40.0), 
hmap(newdata, compData, 20, 13, "Δ Largest Conn. Comp.", -5.0, 25.0),
hmap(newdata, compData, 20, 4, "Δ Degree", -5.0, 5.0)
]

plot(fig1hmaps..., colorbar = true, grid=false, layout = (5,2), size = (1000, 1500))

Figure 11 Difference between Figure 9 (with neighborRange) and Figure 4 (neighborMom, neighborRange set to 1) at same inputs, besides findMom method

Heatmap for iterating by neighborRange (2-40) and cl

function dataRead(data, range, index::Int)
    df = CSV.read(data, DataFrame, delim=",")
    df = df[!, [2, 6, 4, 5, 8, 13, 14, 15, 16, 17, 3, 11, 10, 12, 7, 9, 1]] #used for soring by clratio/distFactor/distNeighborRange, including graphCalcs
    sort!(df, [:distNeighborRange, :cl])
    dfData = Matrix{Float64}(df)
    dataArr = zeros(range, range, 1) #transforms [range^2 * 10] arr to [range, range, 1] arr
    for(i) in 1:range
        for(j) in 1:range
            dataArr[i, j, 1] = dfData[range*(i-1)+j, index]
        end
    end
    dataArr
end

function hmap(data::String, range::Int, index::Int, titles::String, cLow::Float64, chI::Float64) 
    #indexes : 1 yidx   2 xidx   3 coopFreq   4 degree   5 fitness  6 pnc_end   7 pnd_end   8 prc_end  9 prd_end  10 shortestPath 11 connComponents 12 meanCCSize 13 largestCC 14 mean connection distance 15 distance  16 inclusion  17 assortment 
    dataArr = dataRead(data, range, index)
    x_axis = String[]
    y_axis = String[]
    for(i) in 1:range 
        push!(y_axis, (string(Int(40*i/range))))
        push!(x_axis, (string(round(0.4*i/range; digits = 3))))
    end
    p = heatmap(x_axis, y_axis,  dataArr[:, :, 1]; title = titles, xlabel="CL",ylabel="Neighbor Range", clim=(cLow,chI))
    p
end

hmap (generic function with 2 methods)

file = "neighborRange-2-40.csv"
fig1hmaps = [hmap(file, 20, 3, "Cooperation Frequency", 0.0, 1.0), 
hmap(file, 20, 6, "PNC", 0.0, 0.8), 
hmap(file, 20, 7, "PND", 0.0, 0.8),
hmap(file, 20, 8, "PR", 0.0, 0.6),
hmap(file, 20, 14, "Mean Dist. of Connections", 1.0, 28.0),
hmap(file, 20, 10, "Mean Shortest Path", 1.0, 6.0), 
hmap(file, 20, 11, "Connected Components", 1.0, 40.0), 
hmap(file, 20, 12, "Mean Conn. Comp. Size", 1.0, 100.0), 
hmap(file, 20, 13, "Largest Conn. Comp.", 1.0, 100.0),
hmap(file, 20, 4, "Degree", 1.0, 35.0)
]

plot(fig1hmaps..., colorbar = true, grid=false, layout = (5,2), size = (1000, 1500))

Figure 12 High strength of selection with the mother selected from individuals a distance neighborRange or below from, pnd but not prd evolving separately. The results above were created with: B=2.0, C=0.5, D=0.0, gen=100000, distInherit=false, findMom=neighborMom, pn=0.5, pnd=true, pr=0.0001, prd=false, muP=0.01, delta=0.5, sigmapn=0.01, sigmapr=0.01, reps=10. Initial cooperation frequency was 0.5.

Finding the difference from both runs with anyMom() and neighborMom() and Figure 12. New heatmap below.

using JLD2
using StatsBase
using Plots
#backend(:plotly)
using DataFrames
using CSV

function benClDataRead(data, range, index::Int)
    df = CSV.read(data, DataFrame, delim=",")
    df = df[!, [3, 2, 5, 6, 8, 13, 14, 15, 16, 17, 4, 11, 10, 12, 7, 9, 1]] #used for soring by ben/cl, including graphCalcs
    #df = df[!, [6, 9, 2, 3, 7, 8, 10, 4, 5, 1]] #used for when sorting by pn/r
    #sort!(df, [:pn, :pr]) #used for when sorting by pn/r
    sort!(df, [:cl, :ben])
    dfData = Matrix{Float64}(df)
    dataArr = zeros(range, range, 1) #transforms [range^2 * 10] arr to [range, range, 1] arr
    for(i) in 1:range
        for(j) in 1:range
            dataArr[i, j, 1] = dfData[range*(i-1)+j, index]
        end
    end
    #print(dataArr)
    dataArr
end

function dataReadMinus(data, data2, range, index::Int)
    df = CSV.read(data, DataFrame, delim=",")
    df = df[!, [2, 6, 4, 5, 8, 13, 14, 15, 16, 17, 3, 11, 10, 12, 7, 9, 1]] #used for soring by clratio/distFactor/distNeighborRange, including graphCalcs
    sort!(df, [:cl, :distNeighborRange])
    dfData = Matrix{Float64}(df)
    #print(df)
    dataArr = zeros(range, range, 1) #transforms [range^2 * 10] arr to [range, range, 1] arr
    for(i) in 1:range
        for(j) in 1:range
            dataArr[i, j, 1] = dfData[range*(i-1)+j, index]
        end
    end
    benclData = benClDataRead(data2, range, index)
    for(i) in 1:range
        dataArr[i, :, 1] .-= benclData[i, 4, 1]
        #dataArr .-= benclData[:, 4, 1]
        #println(dataArr[i, :, 1], " - ", benclData[i, 4, 1])
    end   
    dataArr
end

function hmap(data::String, data2::String, range::Int, index::Int, titles::String, cLow::Float64, chI::Float64) 
    #indexes : 1 yidx   2 xidx   3 coopFreq   4 degree   5 fitness  6 pnc_end   7 pnd_end   8 prc_end  9 prd_end  10 shortestPath 11 connComponents 12 meanCCSize 13 largestCC 14 mean connection distance 15 distance  16 inclusion  17 assortment 
    dataArr = dataReadMinus(data, data2, range, index)
    x_axis = String[]
    y_axis = String[]
    for(i) in 1:range 
        push!(y_axis, (string(Int(40*i/range))))
        push!(x_axis, (string(round(0.4*i/range; digits = 3))))
    end
    p = heatmap(x_axis, y_axis,  dataArr[:, :, 1]; title = titles, xlabel="CL",ylabel="Neighbor Range", clim=(cLow,chI))
    p
end

hmap (generic function with 2 methods)

newdata = "neighborRange-2-40.csv"
compData = "anyMom_pnd_new.csv"

fig1hmaps = [hmap(newdata, compData, 20, 3, "Δ Cooperation Frequency", -0.25, 0.15), 
hmap(newdata, compData, 20, 6, "Δ PNC", -0.3, 0.3), 
hmap(newdata, compData, 20, 7, "Δ PND", -0.3, 0.3),
hmap(newdata, compData, 20, 8, "Δ PR", -0.1, 0.05),
hmap(newdata, compData, 20, 14, "Δ Mean Dist. of Connections", -12.5, 12.5),
hmap(newdata, compData, 20, 10, "Δ Mean Shortest Path", -0.5, 2.0), 
hmap(newdata, compData, 20, 11, "Δ Connected Components", -25.0, 10.0), 
hmap(newdata, compData, 20, 12, "Δ Mean Conn. Comp. Size", -40.0, 30.0), 
hmap(newdata, compData, 20, 13, "Δ Largest Conn. Comp.", -25.0, 35.0),
hmap(newdata, compData, 20, 4, "Δ Degree", -5.0, 5.0)
]

plot(fig1hmaps..., colorbar = true, grid=false, layout = (5,2), size = (1000, 1500))

Figure 13 Difference between Figure 12 (with neighborRange) and Figure 4 (anyMom) at same inputs, besides findMom method

newdata = "neighborRange-2-40.csv"
compData = "neighbor_mom_pnd.csv"

fig1hmaps = [hmap(newdata, compData, 20, 3, "Δ Cooperation Frequency", -0.15, 0.20), 
hmap(newdata, compData, 20, 6, "Δ PNC", -0.3, 0.3), 
hmap(newdata, compData, 20, 7, "Δ PND", -0.3, 0.3),
hmap(newdata, compData, 20, 8, "Δ PR", -0.05, 0.1),
hmap(newdata, compData, 20, 14, "Δ Mean Dist. of Connections", -5.0, 12.5),
hmap(newdata, compData, 20, 10, "Δ Mean Shortest Path", -2.0, 0.5), 
hmap(newdata, compData, 20, 11, "Δ Connected Components", -10.0, 20.0), 
hmap(newdata, compData, 20, 12, "Δ Mean Conn. Comp. Size", -10.0, 40.0), 
hmap(newdata, compData, 20, 13, "Δ Largest Conn. Comp.", -5.0, 25.0),
hmap(newdata, compData, 20, 4, "Δ Degree", -5.0, 5.0)
]

plot(fig1hmaps..., colorbar = true, grid=false, layout = (5,2), size = (1000, 1500))

Figure 14 Difference between Figure 12 (with neighborRange) and Figure 4 (neighborMom, neighborRange set to 1) at same inputs, besides findMom method

Stabilization Testing

The following were testing iterating by number of generations (50000, 75000, 100000, 150000, 250000) and cl, to see when cooperation frequency stabilizes.

Heatmap for Iterating by Gen/CL

function dataRead(data, range, index::Int)
    df = CSV.read(data, DataFrame, delim=",")
    df = df[!, [2, 3, 5, 6, 8, 13, 14, 15, 16, 17, 4, 11, 10, 12, 7, 9, 1]] #used for soring by ben/cl, including graphCalcs
    #df = df[!, [6, 9, 2, 3, 7, 8, 10, 4, 5, 1]] #used for when sorting by pn/r
    #sort!(df, [:pn, :pr]) #used for when sorting by pn/r
    sort!(df, [:ben, :cl])
    dfData = Matrix{Float64}(df)
    dataArr = zeros(range, range, 1) #transforms [range^2 * 10] arr to [range, range, 1] arr
    for(i) in 1:range
        for(j) in 1:range
            dataArr[i, j, 1] = dfData[range*(i-1)+j, index]
        end
    end
    #print(dataArr)
    dataArr
end

function hmap(data::String, range::Int, index::Int, titles::String, cLow::Float64, chI::Float64) 
    #indexes : 1 yidx   2 xidx   3 coopFreq   4 degree   5 fitness  6 pnc_end   7 pnd_end   8 prc_end  9 prd_end  10 shortestPath 11 connComponents 12 meanCCSize 13 largestCC 14 mean connection distance 15 distance  16 inclusion  17 assortment 
    dataArr = dataRead(data, range, index)
    x_axis = String[]
    y_axis = String[]
    gens = [50000, 75000, 100000, 150000, 250000]
    for(i) in 1:range #PNC/PND times 10
        push!(x_axis, (string(round((0.4*i/range); digits = 4))))
        push!(y_axis, (string(gens[i])))
    end
    p = heatmap(x_axis, y_axis,  dataArr[:, :, 1]; title = titles, xlabel="CL",ylabel="Generations", clim=(cLow,chI))
    p
end

#redefined hmap() in between runs

hmap (generic function with 2 methods)

file = "testing-anymom.csv"
fig1hmaps = [hmap(file, 5, 3, "Cooperation Frequency", 0.0, 1.0), 
hmap(file, 5, 6, "PNC", 0.0, 0.8), 
hmap(file, 5, 7, "PND", 0.0, 0.8),
hmap(file, 5, 8, "PR", 0.0, 0.04),
hmap(file, 5, 11, "Connected Components", 1.0, 40.0), 
hmap(file, 5, 12, "Mean Conn. Comp. Size", 1.0, 100.0), 
]
#,

plot(fig1hmaps..., colorbar = true, grid=false, layout = (3,2), size = (1000, 900))

Figure 15 High strength of selection with the mother selected from any individuals, pnd but not prd evolving separately. The results above were created with: B=2.0, C=0.5, D=0.0, CL=pard["cl"], gen=Int(pard["ben"]), findMom=anyMom, distInherit=false, distFactor=0.975, pn=0.5, pnd=true, pr=0.0001, prd=false, muP=0.01, delta=0.5, sigmapn=0.01, sigmapr=0.01, reps=10. Initial cooperation frequency was 0.5

file = "testing-nmom.csv"
fig1hmaps = [hmap(file, 5, 3, "Cooperation Frequency", 0.0, 1.0), 
hmap(file, 5, 6, "PNC", 0.0, 0.9), 
hmap(file, 5, 7, "PND", 0.0, 0.9),
hmap(file, 5, 8, "PR", 0.0, 0.14),
hmap(file, 5, 11, "Connected Components", 1.0, 20.0), 
hmap(file, 5, 12, "Mean Conn. Comp. Size", 1.0, 100.0), 
]
#,

plot(fig1hmaps..., colorbar = true, grid=false, layout = (3,2), size = (1000, 900))

Figure 16 High strength of selection with the mother selected from dead individuals' neighbors, pnd but not prd evolving separately. The results above were created with: B=2.0, C=0.5, D=0.0, CL=pard["cl"], gen=pard["ben"], findMom=neighborMom, distInherit=false, distFactor=0.975, pn=0.5, pnd=true, pr=0.0001, prd=false, muP=0.01, delta=0.5, sigmapn=0.01, sigmapr=0.01, reps=10. Initial cooperation frequency was 0.5 .